Impact of co-adsorbed oxygen on crotonaldehydeadsorption over gold nanoclusters: acomputational study†‡
Constantinos D. Zeinalipour-Yazdi,ab David J. Willock,*b Andreia Machado,b
Karen Wilsonc and Adam F. Lee*ad
Crotonaldehyde (2-butenal) adsorption over gold sub-nanometer particles, and the influence of
co-adsorbed oxygen, has been systematically investigated by computational methods. Using density
functional theory, the adsorption energetics of crotonaldehyde on bare and oxidised gold clusters (Au13,
d = 0.8 nm) were determined as a function of oxygen coverage and coordination geometry. At low
oxygen coverage, sites are available for which crotonaldehyde adsorption is enhanced relative to bare Au
clusters by 10 kJ mol�1. At higher oxygen coverage, crotonaldehyde is forced to adsorb in close proximity
to oxygen weakening adsorption by up to 60 kJ mol�1 relative to bare Au. Bonding geometries, density of
states plots and Bader analysis, are used to elucidate crotonaldehyde bonding to gold nanoparticles in terms
of partial electron transfer from Au to crotonaldehyde, and note that donation to gold from crotonaldehyde
also becomes significant following metal oxidation. At high oxygen coverage we find that all molecular
adsorption sites have a neighbouring, destabilising, oxygen adatom so that despite enhanced donation,
crotonaldehyde adsorption is always weakened by steric interactions. For a larger cluster (Au38, d = 1.1 nm)
crotonaldehyde adsorption is destabilized in this way even at a low oxygen coverage. These findings provide
a quantitative framework to underpin the experimentally observed influence of oxygen on the selective
oxidation of crotyl alcohol to crotonaldehyde over gold and gold–palladium alloys.
I. Introduction
Platinum group metal (PGMs) and gold nanoparticles (NPs) arepromising heterogeneous catalysts for the selective aerobicoxidation (selox) of alcohols to aldehydes, providing routes thatobviate the use of environmentally harmful inorganic oxidants(e.g. CrVI salts1 or permanganates2) and their associated toxicwaste,3 handling of hazardous peroxides, or the recovery ofexpensive organometallic soluble catalysts. Recent research hashighlighted Au,4 Pd5,6 and bimetallic nanoparticles thereof7–10
as atom-efficient selox catalysts, able to operate under mildconditions (e.g. reaction temperatures between 60–160 1C and
employing ambient air as an oxidant) in non-halogenatedsolvents, aqueous solutions11 or even solventless.7 The activesite in oxide supported Pd NPs for crotyl alcohol selox has beenextensively investigated by in situ and operando X-ray spectro-scopies.12–15 These reveal that oxygen is critical in weakeningthe adsorption of the reactively-formed crotonaldehyde productthereby reducing decarbonylation pathways. We have madesimilar observations over Pd(111)16 and Au/Pd(111)9,17 modelsingle-crystal catalysts, wherein in situ XPS and TPRS measure-ments show that pre-adsorbed oxygen weakens crotonaldehydeadsorption over pure palladium and Au/Pd surface alloys, thussuppressing undesired decarbonylation pathways. This contrastswith epoxidation wherein oxygen promotes ethene adsorptionover silver.18 Enhanced activity of Au/Pd core shell nanoparticlesin crotyl alcohol selox has been rationalized by d-charge depletionobserved via XANES measurements.10
Chin et al. recently explored the influence of co-adsorbedoxygen on methane combustion over Pd and Pt NPs,19,20 whileethanol selective oxidation to ethanal was studied over Au NPsvia DFT calculations by Boronat and Corma,21 who found thatoxygen lowered the activation barriers for O–H and C–H scission.Delbecq and Sautet also employed DFT in combination withHREELS22 vibrational measurements to study crotonaldehyde
a Department of Chemistry, University of Warwick, Coventry, CV4 7AL, UK.
E-mail: [email protected] Cardiff Catalysis Institute, School of Chemistry, Cardiff University,
Cardiff CF10 3AT, UK. E-mail: [email protected] European Bioenergy Research Institute, Aston University, Birmingham B4 7ET, UKd School of Chemistry, Monash University, Victoria 3800, Australia
† Electronic supplementary information (ESI) available: The methodology andunit cell for crotonaldehyde adsorption on Au(111), the optimized structures of allnanoparticle/adsorbate models and LDOS plots are given. See DOI: 10.1039/c3cp53691b‡ This Article is Published in Celebration of the 50th Anniversary of the Openingof the Chemistry Department at the University of York.
Received 1st September 2013,Accepted 21st November 2013
adsorption over Pt(111)23 and Pt2Sn/Pt(111)24 alloys, and therebyunderstand the influence of Sn additives on catalyst selectivitytowards the hydrogenation of crotonaldehyde and other a-b-unsaturated aldehydes. DFT has also been applied in conjunctionwith metastable de-excitation spectroscopy to elucidate the mole-cular adsorption modes and reaction pathways during crotylalcohol selox over bare Pd(111).25
In this contribution, the effect of co-adsorbed oxygen on theadsorption of crotonaldehyde is explored for the first time viaDFT calculations. In particular, we provide fundamental insightinto the impact of co-adsorbed oxygen upon crotonaldehydeadsorption over Au NPs, and discuss the resulting implicationsfor crotyl alcohol selox. We present a systematic computationalstudy of crotonaldehyde adsorption over gold NPs (diameter,d o 1 nm) as a function of co-adsorbed oxygen surface coverageand proximity. The configurational space for adsorbed croton-aldehyde on bare Au13 and Au38 NPs and their oxidised analo-gues (Au13On, n = 2, 4, 6 and 8 and Au38O2) at effective oxygencoverages between 0–1 monolayers has been explored, andassociated adsorption energies calculated. For low oxygen cover-age on Au13 particles we find that crotonaldehyde adsorption canbe enhanced relative to the bare Au cluster. However, strongdestabilisation of crotonaldehyde by co-adsorbed oxygen isfound at full oxygen coverage because adsorbed oxygen adatomsoccupy positions close to the molecular adsorption site. Theconcepts discussed are likely extendable to other PGM NPs, andsuggest that controlling oxygen surface coverage in situ may aidprocess optimization during the catalytic selox of allylic alcohols.
II. Methods
G point26 DFT calculations were performed with the VASP 5.2code.27,28 Exchange and correlation effects are consideredwithin the generalized gradient approximation (GGA) usingthe Perdew–Burke–Ernzerhof (PBE) exchange–correlation (XC)functional,29 with the projector augmented-wave (PAW) method30,31
used to represent core states, 1s for C and O, and 1s to 4f for Au.The cut-off energy for the plane-wave basis was set to 400 eV.The nanoparticle/adsorbate models were centered within a25 � 25 � 25 Å periodic box to ensure a vacuum gap of 15–20 Åaround the clusters, with most calculated results obtained at thelatter gap. Geometry optimizations of all the atoms in the cluster–adsorbate system were performed with a residual force thresholdof 0.015 eV Å�1 using the conjugate-gradient algorithm. Theconvergence criterion for electronic relaxation was 10�4 eV. Theinitial charge density was obtained by superposition of atomiccharges, and the projection operators were evaluated in reciprocalspace. Spin polarization has been tested in several models butfound to be unimportant, with calculated adsorption energiesagreeing to within 0.5 kJ mol�1 for calculations with andwithout spin polarisation.
Molecular gas phase optimizations were also performed with theGaussian09 code (rev C.01)32 to allow comparison of results with theBecke’s three-parameter, hybrid exchange functional33 combinedwith the Lee–Yang–Parr non-local correlation functional,34
abbreviated as B3LYP, using the aug-cc-pVTZ35 basis set. Lineardependencies of the basis functions were removed by using thespherical version (5d, 7f) of the aug-cc-pVTZ basis set.
Adsorption energies (Eads) were calculated by subtracting thetotal energy of the fully-relaxed, isolated adsorbate and nanoparticlefrom the total energy of the adsorbate-nanoparticle system.
The degree of adsorption-induced charge transfer was esti-mated through the atomic charges calculated by Bader’s atoms-in-molecules method using the numerical grid based approachdeveloped by Henkelman and co-workers.36,37 These authorshave noted that the convergence of Bader charges requires thecharge density to be produced on a fine grid so that the spacingbetween grid points is of the order of 0.02 Å.37 VASP works withtwo grids to represent the charge density, and outputs the finergrid for analysis. The memory requirement for such calculationson the 25 Å cube used to isolate the clusters within the periodicboundary conditions of the simulation was prohibitive on ourresources. Accordingly, charge analysis was carried out for eachsystem within a 16 Å cubic simulation cell using the structuresoptimized within the larger cell. This allowed us to produce agrid with 700 points on a side corresponding to a grid spacing of0.023 Å. As a check on the numerical accuracy of the analysis wecarried out additional calculations for ethene adsorbed in a di-sfashion to the Au13 cluster. In this case Bader analysis gave Catom charges of �0.1664 |e| and �0.1640 |e|. Since the carbonatoms for ethene should be equivalent by symmetry we infer anumerical accuracy of 0.005 |e|. Accordingly, Bader charges arequoted to two decimal places throughout the manuscript. Theseethene calculations help inform on CQC double bond inter-actions with the various clusters studied, and provide an addi-tional reference point for discussing the crotonaldehyde results.
III. Results and discussion
The rotational isomers of free crotonaldehyde were first optimizedto determine the lowest energy gas phase conformers. We foundfour local minima on the potential energy surface, denoted asE-(s)-trans, E-(s)-cis, Z-(s)-trans and Z-(s)-cis as depicted in Fig. 1a–d.The heavy atom molecular framework of all rotamers was found tobe planar, indicative of p-conjugation of the CQC and CQOmoieties. As expected for rotational isomers, there is a negligiblechange of the corresponding bond lengths between the variousisomers. The ordering of the relative energies for the fourrotamers was found to be E-(s)-trans (�16.0 kJ mol�1) oE-(s)-cis (�6.6 kJ mol�1) o Z-(s)-trans (�4.2 kJ mol�1) o Z-(s)-cis(0.0 kJ mol�1). Of the four rotational isomers identified, E-(s)-trans-crotonaldehyde was the most stable in the gas phase, in agree-ment with microwave spectroscopy experiments38 on trans-crotonaldehyde. It is noteworthy that this energy trend is found forboth plane-wave and atom-centered basis (i.e. B3LYP/aug-cc-pVTZ)methodologies. Subsequently we have used E-(s)-trans to evaluatethe effect of co-adsorbed oxygen on the adsorption strength of thedesirable product in the selective oxidation of crotyl alcohol,although the general conclusions that follow have also beenvalidated for the second lowest energy conformer (E-(s)-cis).
For the gold nanoparticle two particle sizes have been used,Au13 and Au38. In each case the structures were taken from bulkgold (a = b = g = 901, a = b = c = 407.82 pm)39 and fully optimizedto their nearest energetic minima. Both Au13 and Au38 NPsexpose (100)-like and (111)-like facets which are also found on goldNPs of different geometries. Au38 has previously been described asan efficient catalyst for molecular oxygen dissociation with DFTcalculation results being confirmed experimentally.40,41 Afteradsorption of molecular oxygen on a (100) facet a low barrierwas calculated to produce O-adatoms in 3-fold hollow sites (h-O)on adjacent (111) sections of the particle. In Fig. 2 we considerthe energetics of h-O adsorption on an Au13 cluster by plottingthe relative energy of the oxidised cluster compared to Au13 and thecorresponding number of gas phase O2 molecules as a function ofoxygen coverage. Successive addition of oxygen adatoms thermo-dynamically stabilizes this system due to the formation oflinear O–Au–O structures, which have been previously reportedduring O2 dissociation on gold clusters42 and employed as anindicator of gold oxide formation.43 In particular, we find the
relative energy calculated for oxidised Au13 is linearly depen-dent on the oxygen coverage. This suggests that in an oxidisingatmosphere the Au nanoparticles considered here will becomesurface oxidised.
We note that DFT in general suggests that small isolatedclusters of bare gold (e.g. Au4,44 Au12
45) preferentially adopt aplanar structure, whereas larger Aun clusters (n 4 13) favour 3Dmorphologies that introduce low-coordination edge and vertexsites alongside (100) facets. However, for surface supportednanoparticles, three-dimensional morphologies46,47 are observedeven for small particles. In the present context we note thatoxidation of nanoparticles leads to linear O–Au–O structures42 asAu atoms are oxidized,43 and that these will for thermodynamicreasons more easily be accommodated within three dimensionalstructures.48 Accordingly, our use of bulk like structures allowsthe comparison of crotonaldehyde adsorption as a function ofcluster size and oxygen coverage using a consistent clustergeometry which is also in line with the shapes of the supportedAu nano-particles used in selox experimentally.
In the subsequent section we explore the various adsorbedconfigurations of E-(s)-trans-crotonaldehyde on nanometerscale Au particles and examine the effect of co-adsorbed oxygenand cluster size on the adsorption energetics. The four lowestenergy bound structures for each cluster studied are given inFig. 3. For the bare Au13 cluster (Fig. 3a) the most favourableadsorption geometry is pCC which is some 15 kJ mol�1 lower inenergy than the atop and di-rCC modes. The di-rCC structures
Fig. 1 Structure and nomenclature of rotational isomers of crotonaldehydeobtained at Ecut = 400 eV in a 25 � 25 � 25 Å periodic cubic cell. All bondlengths given in Å.
Fig. 2 Relative total energy of Au13On, where n = 0, 2, 4, 6 and 8 for thevarious possible permutations of three-fold hollow oxygen. The least squaresline is drawn for minimum energy structure at each oxygen coverage.
Fig. 3 Adsorption energies of di-rCC, pCC and atop-bonded conforma-tions of E-(s)-trans-crotonaldehyde on neutral (a) Au13, (b) Au13(h-O)8, (c)Au38 and (d) Au38(h-O)2.
are akin to ethylene adsorption on neutral, anionic and cationicgold clusters.49 For E-(s)-trans crotonaldehyde adsorption tothe clusters we distinguish two di-rCC geometries; in di-rCC-1the aldehyde group is toward an Au–Au bridge site while fordi-rCC-2 it is orientated away from the cluster. For Au13 andAu38 (Fig. 3c) these two alternatives are practically isoenergetic.However, for the oxidized clusters (Au13(h-O)8, Fig. 3b andAu38(h-O)2, Fig. 3d), locating the aldehyde over the bridgesite places this group near an h-O atom, which is moredestabilizing than locating the methyl group in the sameposition such that di-rCC-2 becomes notably lower in energythat di-rCC-1.
Delbecq and Sautet have provided a survey of adsorptionconfigurations for acrolein (propenal) and crotonaldehyde onPt(111),23,24 along with calculated adsorption energies usingthe PW91 functional. For the E-(s)-trans isomer they consideredfive adsorption modes. Two of these adsorb via the CQC bond;di-rCC for which CQC interacts with two surface Pt atoms andpCC which involves only a single Pt atom. Adsorption via theCQO moiety gives a further two possibilities; di-rCO and atopwith the molecule adsorbed in an end-on fashion through aPt� � �OQC interaction. Finally, a Z4 configuration was examinedinvolving both CQC and CQO bonding to separate Pt atoms.For crotonaldehyde this latter Z4 adsorption geometry gave thelowest calculated adsorption energy (�84.5 kJ mol�1).
The adsorption configurations we have located are in generalagreement with the earlier work on Pt. Interestingly, the Z4
structure for adsorbed E-(s)-trans crotonaldehyde previouslyshown to have the potential to undergo decarbonylation overPd(111)25 was not found on Au13O8. This implies that oxidisedsub-nanometer Au NPs may afford higher crotonaldehydeselectivity rather than decarbonylation, during crotyl alcoholto crotonaldehyde selox. For the remaining configurations,we find that adsorption through the CQC bond and CQOmoieties are possible on both Au13 and Au38 clusters. The NPstructure gives rise to edge and corner sites where the facetsmeet and these are found to be preferred over adsorption onthe facets themselves. For Au13 structures bound through theCQC moiety generally adsorb more strongly than those viaoxygen bound atop (Fig. 3a). However, in the presence ofco-adsorbed oxygen we find that atop adsorption actuallybecomes more favourable (Fig. 3b). Side reactions that couldpotentially lead to attack at the allylic bond in the CQC boundconfigurations di-rCC-1, di-rCC-2 and pCC are thus less probablefor oxidized sub-nanometer Au NPs because atop adsorptionthrough the aldehyde oxygen atom is energetically favoured.For Au38 this preference is less clear as the different adsorptionmodes shown in Fig. 3c fall within a narrow range (3.6 kJ mol�1).We have also examined the adsorption energy of di-s andp-bound E-(s)-trans-crotonaldehyde on Au(111) (see ESI,†Fig. S1) and found that the adsorption energy is essentiallyzero (1.1 kJ mol�1) indicative that on extended gold surfacescrotonaldehyde can only weakly interact. Comparing thelowest adsorption energies for Pt(111),22,23 Au NPs andAu(111) yields the following trend for crotonaldehyde:Pt(111) (Z4: �85 kJ mol�1) 4 Au13 (pCC: �71 kJ mol�1) 4 Au38
It is remarkable that the smaller Au cluster gives an adsorp-tion energy just 14 kJ mol�1 less negative than the Pt(111) value.In contrast, the Au38 value is 38 kJ mol�1 less favourable thanPt(111) whereas on Au(111) there is practically no interaction atthe PBE-level, revealing a strong particle size dependence of theadsorption energy. This weakened adsorption would imply apronounced particle size-sensitivity for allylic alcohol selox overgold NPs, with the desired crotonaldehyde product preferentiallydestabilized over larger particles or small NPs that are oxidised(i.e. Au13O8). It is also apparent from Fig. 3 that the energydifferences between adsorption to the different sites on a particlefall into a much narrower range for Au38 (3.1 kJ mol�1) than forAu13 (15.3 kJ mol�1).
Co-adsorbed oxygen exerts an even greater influence oncrotonaldehyde adsorption energetics than does particle size.Specifically, we observe a 94%, 86%, 71% and 51% decrease inthe magnitude of adsorption energies for E-(s)-trans in thedi-rCC-1, di-rCC-2, pCC and atop geometries respectively whenAu13 NPs are pre-saturated with oxygen adatoms; this equates toan average decrease of 45 kJ mol�1 upon nanoparticle oxidation.
In later sections we will show that for Au13 the weakening ofadsorption observed for the di-rCC geometries on the fullyoxidised cluster, Au13(h-O)8, occurs due to the proximity of ah-O atom at the position closest to the bridge site where themolecule is placed (see asterisk in Fig. 3a). A similar arrangementwas constructed for the partially oxidized Au38(h-O)2 cluster, how-ever, attempts to locate an oxygen atom in the 3-fold hollow siteneighboring the CQC bond destabilized E-(s)-trans-crotonaldehydeto such an extent that no optimized adsorption geometry could befound. Interestingly we find that if we place an oxygen at theneighbouring 3-fold hollow and keep di-s-bound crotonaldehydeat a fixed position, then oxygen diffusion to the adjacent (111)-likefacet is observed. However, for Au38(h-O)2, the presence of oxygenadatoms at other adjacent sites also decreased the magnitude ofcrotonaldehyde adsorption by between 10–45 kJ mol�1 (compareFig. 3c and d). Such findings provide evidence for a strong oxygen-induced destabilization of allylic aldehyde adsorption over goldnanoparticles. The origin of this dramatic destabilization, and itsoxygen coverage dependence, reflects an intriguing and complexbonding mechanism that we explore in the following sections.
Boronat and Corma have shown that molecular O2 adsorp-tion on the (100) facets of Au38 clusters can yield h-O adatomson neighbouring (111) facets. We have used the idea of oxygendissociation on (100) leading to h-O on opposing (111) facets tobuild-up the oxygen adatom population on an Au13 cluster andexamined the effect upon E-(s)-trans crotonaldehyde adsorption inthe di-rCC-2 and pCC configurations. The dependence of croton-aldehyde adsorption over Au13 on co-adsorbed oxygen coverage isshown in Fig. 4. In Fig. 4, each structure has an O atom placed inthe hollow site closest to the adsorbate; this clearly has a dramaticeffect on the calculated adsorption energy even at low coverage.Indeed for the di-rCC configuration, dissociation of a single oxygenmolecule into such an arrangement leads to 53.1 kJ mol�1 fall in
the crotonaldehyde adsorption energy; for intermediate cover-ages, progressive addition of further oxygen adatoms slightlyenhances adsorption with calculated energies increasing from�2.1 (Au13(h-O)2) to �16.5 kJ mol�1 (Au13(h-O)6).
At full coverage (Au13(h-O)8) adsorption weakens again,being destabilized by 47.2 kJ mol�1 with respect to the bareAu13 cluster. A very similar trend is also observed for the pCC
configuration, where on average the adsorption energies are11 kJ mol�1 stronger than for the di-rCC configuration (exclud-ing Au13O2). For this crotonaldehyde adsorption geometry overAu13O2, the adjacent h-O shifts away from the aldehyde toa two-fold bridged oxygen (b-O) position, relieving the stericinteraction.
Fig. 4 highlights the impact of h-O oxygen atoms located inthe (111)-like adsorption site neighboring an adsorbed croton-aldehyde molecule. Such proximate adsorption to the CQCbond dramatically reduces the strength of crotonaldehydeadsorption. However, there are clearly other sites available forall intermediate oxygen coverage levels, which are only lost at thepoint of hollow site saturation (Au13(h-O)8). Fig. 5 explores theeffect of h-O oxygen atom placement on the adsorption of E-(s)-trans crotonaldehyde in the di-rCC-2 geometry over Au13(h-O)6.The results fall into two broad categories. When the proximalh-O site is occupied, crotonaldehyde adsorption is destabilized(as reported in Fig. 4). Fig. 5 reveals that the oxygen adatomarrangement in Fig. 4 actually represents one of the morestrongly bound examples of such structures, with crotonaldehydeadsorption varying between +0.8 and�16.5 kJ mol�1. Locating anoxygen adatom in the (111)-like hollow site nearest to theadsorbed molecule weakens crotonaldehyde binding by an aver-age of 59 kJ mol�1. By comparison, when the proximal h-O site isunoccupied, crotonaldehyde adsorption is enhanced by between5.5 and 13.6 kJ mol�1 relative to the bare Au13 cluster. Thisfinding has clear implications for practical catalysis, since occu-pancy of the proximal h-O site can only be guaranteed atsaturation oxygen adatom surface coverages, with high binding
energy sites for strong crotonaldehyde adsorption available onpartially oxidized gold NPs. High oxygen partial pressures/dissolved concentrations should thus eliminate reaction path-ways involving crotonaldehyde adsorption at strongly adsorbingsites which likely favour decarbonylation versus desired compet-ing desorption pathways. Fully oxidized gold NP catalysts areanticipated to disfavour the adsorption of related allylic alde-hydes with respect to their naked gold counterparts, and mayalso interact more weakly with other unsaturated organic mole-cules that typically coordinate through CQC functions.
Of the two clusters explored in the present work, Au13 is uniquein that each (111) like facets contains only a single three-foldhollow site. Larger clusters possess extended facets, so that evenon Au38 each (111) face contains six hollow sites. Occupancy of allsuch (111) hollow sites by oxygen adatoms is unlikely since thiswould place h-O atoms close to one another. Boronat andCorma21 suggest that a Au38O16 cluster, in which only a third ofthe available h-O sites are occupied, represents a saturated oxidemonolayer. In this latter scenario, crotonaldehyde adsorptioncould away from a neighboring h-O atom. It is noteworthy fromFig. 3 that such open adsorption sites on larger clusters would stillbe destabilized with respect to the bare cluster by 9.7 kJ mol�1
(pCC) and 44.4 kJ mol�1 (di-rCC-1). Hence, nanometre gold nano-particles partially oxygen covered remain able to adsorb croton-aldehyde with moderate energies (�20 to �40 kJ mol�1) at sitesthat are not directly adjacent to oxygen adatoms.
Fig. 5 reveals that oxygen adsorption over for Au13 can eitherenhance or destabilize crotonaldehyde adsorption, and we nowadvance a bonding model to explain these observations. Thedi-rCC mode of bonding can be viewed in terms of donationfrom the CQC p-density to empty metal orbitals, and back-donation from the filled metal d-states into the anti-bondingp*-orbitals of the allylic bond. These combined effects changethe structures of CQC double bonds as the carbon atoms alterfrom sp2 toward sp3 geometries. The structural analyses pre-sented in Fig. 6 for crotonaldehyde over bare and oxidized Au13
NPs reveal qualitative trends in line with this model. We findgenerally shorter di-rCC bond lengths (rAu–C) for the morestrongly adsorbed configurations (Fig. 6a) accompanied by adecreased dihedral angle jH–C2–C3–H for crotonaldehyde, indi-cative of sp2 to sp3 rehybridization of the two C atoms in theolefinic bond. For metals with a filled d-shell such as Au, wewould also expect the interaction with the bare cluster to bedominated by metal to adsorbate back donation.
The sub-set of structures lacking an h-O atom adjacent tocrotonaldehyde exhibit stronger di-rCC-2 adsorption thanoccurs on the bare Au13 cluster (Fig. 5). These structures havecorrespondingly shorter rAu–C bonds by B0.043 Å, and smallerjH–C2–C3–H dihedral angles (1421 versus 1471) than for the barecluster, consistent with strong rehybridization of the two Catoms in the CQC moiety in the presence of h-O adatoms. Thislikely reflects oxidation of the Au atoms to which the croton-aldehyde is bound; cationic gold, having a smaller effectiveradius than the neutral atom species, permits closer approachof the adsorbate, while the associated depopulation of d-statesfacilitates adsorbate to cluster charge donation in concert with
Fig. 4 Adsorption energies of di-rCC-2 (black lines) and pCC (grey lines)bound E-(s)-trans-crotonaldehyde as a function of oxygen coverage onneutral Au13, with an O adatom positioned in the hollow site proximate tothe molecule.
the metal to adsorbate back donation that also operates on thebare cluster. The two Au atoms coordinating to crotonaldehydeare also 0.046 Å closer in this set of oxidized clusters than theyare for the parent Au13 cluster, probably reflecting distortion ofthe former arising from the stronger CQC binding.
A similar picture emerges for the sub-set of structures inwhich the three-fold hollow site closest to crotonaldehyde isoccupied by oxygen, for which aldehyde adsorption is signifi-cantly weaker than on bare Au13. These configurations havegenerally longer rAu–C bonds (spanning 2.156 Å to 2.222 Å) andjH–C2–C3–H dihedral angles closer to the 1801 value expected forthe planar CQC double bond in the unbound aldehyde.Although the two coordinating Au atoms in this sub-set areagain expected to be electron deficient due to adsorbed oxygen,the attractive interactions from molecule-to-metal electrondonation have to be balanced against the steric repulsionbetween crotonaldehyde and the proximate h-O oxygen atom.Fig. 6c shows that oxygen adatoms placed near crotonaldehydesignificantly increase the separation of Au atoms to which theCQC bond coordinates, resulting in an inverse correlationbetween aldehyde adsorption and Au–Au bond distance.
A simple trigonometric calculation shows that, for sp3 hybridisedC atoms, an Au–Au spacing of around 2.80 Å is optimal forAu–C distances of 2.16 Å, i.e. the most strongly bound clustershave geometries consistent with this bonding geometry at theC atoms.
Fig. 7 shows a comparison of the Au 5d partial density ofstates (PDOS) for the Au13 cluster and associated Au13O6
structures. The bare Au13 states range from around �7 to�1 eV relative to the Fermi level, consistent with the filledd-shell of atomic Au, whereas the Au13O6 PDOS plots show abroader spread of energies, with unoccupied states presentabove the Fermi level reflecting the formation of linearO–Au–O bonds and concomitant gold oxidation. Formation ofempty metal d-states upon cluster oxidation permits localizedadsorbate - metal donation as discussed above.
Further insight into the charge distribution of selectedconfigurations was obtained using Bader analysis (Table 1).Fig. 7 shows that the Au13 cluster has filled d-orbitals and so wewould expect a net electron donation from the d-orbitals of thebare Au13 cluster to adsorbed crotonaldehyde with negligibleback-donation, however only minimal charge transfer (Dq = 0.04)
Fig. 6 Relationship between crotonaldehyde adsorption energy DEads and (a) average gold–carbon bond length rAu–C between crotonaldehyde andAu13 cluster; (b) dihedral bond of crotonaldehyde jH–C2–C3–H; and (c) the gold–gold bond length rAu–Au following di-rCC adsorption of E-(s)-transcrotonaldehyde. Analysis of geometric parameters was performed on the bare and oxidised Au13 NPs shown in Fig. 5. Note the two cluster–adsorbatesystems with slightly positive adsorption energies in were omitted for clarity.
Fig. 5 Adsorption energies of E-(s)-trans-crotonaldehyde as di-rCC-2 for different arrangements of h-O oxygen atoms on Au13(h-O)6.
is actually calculated. It is worth recalling that gold NPs withodd numbers of electrons, such as the neutral Au13 cluster,exhibit high electron affinities50,51 due to the unpaired electronwhich formally occupies a delocalised state arising from theoverlap of Au sp-atomic orbitals. This high electron affinity mayallow electron transfer from crotonaldehyde to the cluster andTable 1 suggests that (for Au13) these opposing effects self-compensate. Oxygen addition to Au13 to form Au13O6 results ina net charge transfer of 4.39 e from the cluster to oxygenadatoms, with the six Au atoms doubly coordinated to Oa
carrying an average charge of 0.49 e, significantly higher thanthe quoted average across all 13 Au atoms. Crotonaldehydeadsorption to the Au13O6 cluster lacking an h-O adatom neigh-bouring the adsorption site shows significant net electron donationfrom the aldehyde to the cluster (Dq = �0.13), due to the available
empty d-orbitals associated with charge withdrawal by theco-adsorbed oxygen, thereby enhancing crotonaldehyde adsorp-tion relative to Au13 (�68.8 kJ mol�1 versus �55.2 kJ mol�1
respectively). Bader charge analysis for the Au13O6 where a neigh-bouring h-O atom is present, reveals a virtually identical chargedistribution to that of oxidized cluster with this site vacant. Thisimplies that the electronic structure of crotonaldehyde binding toAu13O6 cluster is similar with or without the proximate h-O. Hencethe large energy difference of 52.3 kJ mol�1 between these configu-rations and attendant molecular destabilization associated withthe former can only be attributed to steric interactions between theneighbouring adsorbed oxygen and the aldehyde. Parallel calculationsemploying ethene as a model allylic adsorbate offer a similarpicture; oxidation of the Au13 cluster promotes adsorption unlessthe proximal h-O site is occupied. For ethene, the additionalstabilization due to the availability of empty d-orbitals uponcluster oxidation is greater than for crotonaldehyde (18.2 versus13.6 kJ mol�1). Conversely, ethene destabilization over Au13O6 byneighbouring h-O is lower (30.9 kJ mol�1) than that observed forcrotonaldehyde. The charge distributions for adsorbed ethenewith and without neighbouring h-O are also essentially identical,confirming that for adsorbates possessing CQC bonds, it isrepulsive steric interactions that regulate their adsorption overgold nanoparticles, with ethene less sensitive to such factors thanthe bulkier allylic aldehyde. We note here that the influence ofsurface oxygen atoms on the adsorption of ethene is consideredfor comparison with our crotonaldehyde results. The formation ofa oxametallacycle akin to that discussed by Linic and Barteau forethene epoxidation over Ag52 has not been considered.
Fig. 7 Comparison of partial density of states (PDOS) for the Au 5d statesof Au13 (blue), Au13O6 with h-O at adsorption site (red) and Au13O6 withouth-O at adsorption site (black). The Fermi level, EF, is indicated by the verticaldashed line. A smearing parameter of 0.2 eV was used in the VASP calcula-tions with no additional smearing applied to the resulting PDOS data.
Table 1 Bader charge analysis of crotonaldehyde (CrCHO) and ethene (Et) adsorption to Au13 and Au13O6 clusters
di-r-Et-Au13O6 no neighboring O 4.26 �4.44 0.33 �0.74 �0.18 �0.15 �0.14 — �102.2
di-r-Et-Au13O6 with neighboring O 4.22 �4.43 0.32 �0.74 �0.21 �0.14 �0.13 — �53.1
a Sum of Bader charges on indicated atom type. b Average Bader charge on indicated atom type. c Charge transfer, negative values indicate etransfer from adsorbate to nanoparticle. d Bader charge on carbon atoms of CQC bonded to the cluster, in CrCHO C1 has a CH3 and C2 a CHOsubstituent. e Bader charge on aldehyde oxygen atom.
The electronic effect of the substituents on the croton-aldehyde CQC bond can also be seen; even for the gasphase molecule the two C atoms have significantly differentBader charges. On adsorption both C atoms become slightlymore negatively charged but this effect is in-sensitive tothe state of the cluster. Further the charge on the aldehydeoxygen atom is practically unchanged from the gas phase foreach adsorption mode suggesting that electronic rearrange-ment within the molecule upon adsorption plays a relativelyminor role.
IV. Conclusions
In this study we find that oxidation of well-defined Au nano-particles strongly influences crotonaldehyde and ethene adsorp-tion. For the 0.8 nm diameter Au13 cluster discussed here,intermediate oxygen coverages enhance crotonaldehyde andethene adsorption through the CQC group by 14–18 kJ mol�1.This effect can be understood in terms of the genesis of emptyAu d-states in the oxidised clusters able to receive electronsfrom these allylic adsorbates. At higher oxygen coverages thisbonding picture is still valid, however steric interactions betweenoxygen adatoms neighbouring the crotonaldehyde (ethene)adsorption site destabilize the molecule by up to 60 kJ mol�1
with respect to bare Au13. We also find that oxidation of thesesub-nanometer gold NPs is likely to be favoured in an oxidisingatmosphere, so that under normal selox conditions oxygen willbe present on the surface.
Increasing the gold nanoparticle size to 1.1 nm (Au38) lowersthe crotonaldehyde adsorption energy by 20–40% compared tothe bare Au13 cluster. Oxidation to Au38O2, wherein the O adatomsare located close to co-adsorbed crotonaldehyde, weakens theallylic aldehyde adsorption by as much as 90%.
These calculations highlight the importance of both thesurface coverage, and location of adsorbed oxygen upon thestability of co-adsorbed crotonaldehyde, the desired product ofcrotyl alcohol selox. Oxygen bound in a three-fold hollow siteadjacent to adsorbed crotonaldehyde dramatically destabilizesthe latter due to strong steric repulsion. This finding implies thatcrotyl alcohol selox catalysed over gold nanoparticles should beconducted under high oxygen partial pressures, in order to desorbthe reactively-formed aldehyde product and prevent side-reactionssuch as crotonaldehyde decarbonylation to propene and CO.Conversely, low oxygen partial pressures, and concomitant lowsurface oxygen coverages may hinder crotonaldehyde desorptiondue to enhanced adsorbate - cluster electron transfer relative tounoxidised gold nanoparticles.
Acknowledgements
We thank the EPSRC (EP/E046754/1; EP/G007594/3) forfinancial support and the award of a Leadership Fellowship(AFL) and studentship (AM). Computational resources forthis project were partially provided by UK’s National high-performance computing service, HECToR (EP/F067496) through
the materials consortium, ARCCA and HPC-Wales super-computer facilities. We thank Dr Adam Thetford and SoonWen Hoh for discussions on the charge analysis carried out aspart of this work.
Notes and references
1 D. G. Lee and U. A. Spitzer, J. Org. Chem., 1970, 35, 3589.2 P. V. Prabhakaran, S. Venkatachalam and K. N. Ninian,
Eur. Polym. J., 1999, 35, 1743.3 J. Muzart, Tetrahedron, 2003, 59, 5789.4 C. Della Pina, E. Faletta, L. Prati and M. Rossi, Chem. Soc.
Rev., 2008, 37, 2077.5 K. Mori, K. Yamaguchi, T. Hara, T. Mizugaki, K. Ebitani and
K. Kaneda, J. Am. Chem. Soc., 2002, 124, 11572.6 S. F. J. Hackett, R. M. Brydson, M. H. Gass, I. Harvey, A. D.
Newman, K. Wilson and A. F. Lee, Angew. Chem., Int. Ed.,2007, 46, 8593.
7 D. I. Enache, J. K. Edwards, P. Landon, B. Solsona-Espriu,A. F. Carley, A. A. Herzing, M. Wantanabe, C. J. Kiely,D. W. Knight and G. Hutchings, Science, 2006, 311, 362.
8 D. I. Enache, D. W. Knight and G. Hutchings, Catal. Lett.,2005, 103, 43.
9 A. F. Lee, S. F. J. Hackett, G. J. Hutchings, S. Lizzit,J. Naugthon and K. Wilson, Catal. Today, 2009, 145, 251.
10 T. Balcha, J. R. Strobl, C. Fowler, P. Dash and R. W. J. Scott,ACS Catal., 2011, 1, 425.
11 H. Tsunoyama, H. Sakurai, Y. Negishi and T. Tsukuda,J. Am. Chem. Soc., 2005, 127, 9374.
12 A. F. Lee, C. V. Ellis, J. N. Naughton, M. A. Newton, C. M. A.Parlett and K. Wilson, J. Am. Chem. Soc., 2011, 133, 5724.
13 C. V. Gaskell, C. M. A. Parlett, M. A. Newton, K. Wilson andA. F. Lee, ACS Catal., 2012, 2, 2242.
14 A. F. Lee, J. N. Naughton, Z. Liu and K. Wilson, ACS Catal.,2012, 2, 2235.
15 C. M. A. Parlett, C. V. Gaskell, J. N. Naughton, M. A. Newton,K. Wilson and A. F. Lee, Catal. Today, 2013, 205, 76.
16 A. F. Lee, Z. Chang, P. Ellis, S. F. J. Hackett and K. Wilson,J. Phys. Chem. C, 2007, 111, 18844.
17 J. Naughton, A. F. Lee, S. Thompson, C. P. Vinod andK. Wilson, Phys. Chem. Chem. Phys., 2010, 12, 2670.
18 C. Stegelmann, N. C. Schiodt, C. T. Campbell and P. Stoltze,J. Catal., 2004, 221, 630.
19 Y.-H. Chin and E. Iglesia, J. Phys. Chem. C, 2011, 115, 17845.20 Y.-H. Chin, C. Buda, M. Neurock and E. Iglesia, J. Am. Chem.
Soc., 2011, 133, 15958.21 M. Boronat and A. Corma, J. Catal., 2011, 284, 138.22 J. Haubrich, D. Loffreda, F. Delbecq, Y. Jugnet, P. Sautet,
A. Krupski, C. Becker and K. Wandelt, Chem. Phys. Lett.,2006, 433, 188.
23 F. Delbecq and P. Sautet, J. Catal., 2002, 211, 398.24 F. Delbecq and P. Sautet, J. Catal., 2003, 220, 115.25 J. Naughton, A. Pratt, C. W. Woffinden, C. Eames, S. P. Tear,
S. M. Thompson, A. F. Lee and K. Wilson, J. Phys. Chem. C,2011, 115, 25290.
26 H. J. Monkhorst and J. D. Pack, Phys. Rev. B: Solid State,1976, 13, 5188.
27 G. Kresse and J. Furthmuller, Phys. Rev. B: Condens. MatterMater. Phys., 1996, 54, 11169.
28 G. Kresse and J. Hafner, Phys. Rev. B: Condens. Matter Mater.Phys., 1993, 47, 558.
29 J. P. Perdew, K. Burke and M. Ernzerhof, Phys. Rev. Lett.,1996, 77, 3865.
30 G. Kresse and D. Joubert, Phys Rev. B, 1999, 59, 1758.31 P. E. Blochl, Phys. Rev. B: Condens. Matter Mater. Phys., 1994,
50, 17953.32 M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria,
M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone,B. Mennucci, H. Nakatsuji, G. A. Petersson, M. Caricato,X. Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng,J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda,J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao,T. Vreven, H. Nakai, J. A. Montgomery Jr., J. E. Peralta,F. Ogliaro, M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin,V. N. Staroverov, R. Kobayashi, J. Normand, K. Raghavachari,A. Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi,N. Rega, J. M. Millam, M. Klene, J. E. Knox, V. Bakken, J. B.Cross, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann,O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski,K. Morokuma, R. L. Martin, V. G. Zakrzewski, G. A. Voth,P. Salvador, J. J. Dannenberg, S. Dapprich, A. D. Daniels,O. Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski and D. J.Fox, R. C. Gaussian 09, Gaussian, Inc., Wallingford, CT, 2009.
33 A. D. Becke, J. Chem. Phys., 1993, 98, 5648.34 C. T. Lee, W. T. Yang and R. G. Parr, Phys. Rev. B: Condens.
Matter Mater. Phys., 1988, 37, 785.35 K. A. Peterson, D. E. Woon and T. H. Dunning Jr., J. Chem.
Phys., 1994, 100, 7410.
36 R. F. W. Bader, Atoms in Molecules - A Quantum Theory,Oxford University Press, Oxford, 1990.
37 W. Tang, E. Sanville and G. Henkelman, J. Phys.: Condens.Matter, 2009, 21, 084204.
38 M. Suzuki and K. Kozima, Bull. Chem. Soc. Jpn., 1969,42, 2183.
39 A. Maeland and T. B. Flanagan, Can. J. Phys., 1964, 42,2364.
40 A. Roldan, S. Gonzalez, J. M. Ricart and F. Illas, ChemPhysChem,2009, 10, 348.
41 L. Alves, B. Ballesteros, M. Boronat, J. R. Cabrero-Antonino,P. Concepcion, A. Corma, M. A. Correa-Duarte and E. Mendoza,J. Am. Chem. Soc., 2011, 133, 10251.
42 A. Franceschetti, S. J. Pennycook and S. T. Pantelides,Chem. Phys. Lett., 2003, 374, 471.
43 K. L. Howard and D. J. Willock, Faraday Discuss., 2011,152, 135.
44 C. D. Zeinalipour-Yazdi, A. L. Cooksy and A. M. Efstathiou,Surf. Sci., 2008, 602, 1858.
45 D. F. Mukhamedzyanova, N. K. Ratmanova, D. A. Pichuginaand N. E. Kuz’menko, J. Phys. Chem. C, 2012, 116, 11507.
46 A. A. Herzing, C. J. Kiely, A. F. Carley, P. Landon andG. J. Hutchings, Science, 2008, 321, 1331.
47 M. S. Chen and D. W. Goodman, Science, 2004, 306, 252.48 M. Amft, B. Johansson and N. V. Skorodumova, J. Chem.
Phys., 2012, 136, 024312.49 A. Lyalin and T. Taketsugu, J. Phys. Chem. C, 2009, 114,
2484.50 K. J. Taylor, C. L. Pettiette-Hall, O. Cheshnovsky and R. E.
Smalley, J. Chem. Phys., 1992, 96, 3319.51 H. Hakkinan and U. Landman, Phys. Rev. B: Condens. Matter
Mater. Phys., 2000, 62, R2287.52 S. Linic and M. A. Barteau, J. Am. Chem. Soc., 2002, 124, 310.
